Remediation of soils contaminated by persistent organic pollutants like Polycyclic Aromatic Hydrocarbons (PAHs) has attracted attention after the discovery of thousands of contaminated sites all over the world, due to underground storage leakage or industrial waste disposal. PAHs are considered as persistent, relatively immobile in soils, toxic, mutagenic and carcinogenic. Advanced oxidation process (AOPs) have been proposed in recent years as an attractive alternative for the treatment of matrices contaminated by toxic, refractory or biologically resistent substances. In this work we studied two AOPs: Fenton's reagent and potassium permanganate. Fenton's reagent one of the more typical advanced oxidation processes is based on the property of hydrogen peroxide to generate hydroxyl radicals (OH•) by reacting with ferrous ions and have the potential for rapidly treating or pretreating soils contaminated with toxic and refractory organic wastes. Potassium permanganate is a strong oxidizing agent and has been preferred as an oxidant over ozone and peroxide due to its resistance to auto decomposition and its effectiveness over a larger pH. Permanganate has had success for oxidation of chlorinated solvents for remediation of aquifers and soil. The objective of this work was to: compare the chemical oxidation processes, treatment efficiency of KMnO4 with H2O2 (endogenous iron - Fenton's like reagent or with amendments – Fenton's reagent) for degradation of phenanthrene and pyrene in a French soil and the competitive examination between phenanthrene and pyrene. Soil samples containing 700 mg kg-1 phenanthrene, 1400 mg kg-1 pyrene and total PAHs 2 100 mgkg-1 were the substrates of this work. For Fenton's reagent a fractional factorial design 23-1 was used to examine the effects of three variables: hydrogen peroxide concentration (270-540mmol); iron catalyst concentration (Fe 2+ 27 mmol – endogenous iron) and reaction time (48-72h) on the degradation of the phenanthrene and pyrene. The experiments were carried out in beakers containing 5g of contaminated soil with 10ml water (slurry) to maximize contact of the soil with the liquid. Liquid and solid phases were separated, extracted and analyzed individually to complete the mass balance. In relation to the study with permanganate a factorial design 22 was employed: reaction time (36-72h) and oxidant loading (30-40%) KMnO4/contaminant. The best degradation for phenanthrene, pyrene and PAHs by process Fenton's were as follows: 98.1%-12.1mg kg-1 phenanthrene; 95.6% - 75 mg kg-1 pyrene and 96.5% - 75 mg kg-1 PAHs (540 mmol H2O2; Fe2+ 27 mmol; 72h). The same optimal conditions were determined for three contaminants. In these experiments, iron addition increased treatment efficiency. For treatment with permanganate the best conditions were: 98.6% - 10 mg kg-1 phenanthrene; 96.5% - 49.5 mg kg-1 pyrene and 97.2% - 59.6 mg kg-1 PAHs (30% KMnO4/contaminant; 72h). In these experimental conditions, KMnO4 and Fenton's reagent proved very efficient in the remediation of the phenanthrene, pyrene and PAHs contaminated soils. This efficiency increased with the oxidant loading rate and reaction time. Other, more economical experimental conditions for in situ remediation and ex situ are currently being investigated.